Mailing-List: contact dev-help@hive.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@hive.apache.org
Date: Fri, 24 Oct 2014 13:10:33 +0000 (UTC)
From: "Aidan Semple (JIRA)" <jira@apache.org>
To: hive-dev@hadoop.apache.org
Message-ID: <JIRA.12750343.1414156204000.330629.1414156233607@Atlassian.JIRA>
In-Reply-To: <JIRA.12750343.1414156204000@Atlassian.JIRA>
References: <JIRA.12750343.1414156204000@Atlassian.JIRA>
 <JIRA.12750343.1414156204543@arcas>
Subject: [jira] [Created] (HIVE-8592) 0 values convert to null if casting to
 or inserting to Hive DECIMAL where precision and scale are the same
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Aidan Semple created HIVE-8592:
----------------------------------

             Summary: 0 values convert to null if casting to or inserting to Hive DECIMAL where precision and scale are the same
                 Key: HIVE-8592
                 URL: https://issues.apache.org/jira/browse/HIVE-8592
             Project: Hive
          Issue Type: Bug
          Components: Database/Schema, SQL
    Affects Versions: 0.13.0
         Environment: Running Apache Hive version 0.13.0 using HortonWorks 2.1.2.1 with hadoop version 2.4.0.2.1.2.1-471, on Linux operating system centos5 (also occurs on centos6)
            Reporter: Aidan Semple
             Fix For: 0.13.0


I am trying to load zero values into Hive Decimal fields into a Hive table where the precision and scale are defined as the same e.g. DECIMAL(1,1) or DECIMAL(3,3) etc...
However every time I run a hive ql insert statement to do this containing zero values or run a LOAD DATA command to load a text file of data containing zero values to these columns / fields, on performing a query on the table, these zero values are displayed and treated as NULL values.
On further investigation, I was able to narrow the problem down to doing simple selects with casts. See example and output from Hive below. So attempting to do a cast for 0 or 0.0 or '.0' to DECIMAL(1,1) NULL is returned instead of 0. This is the same for precisions 1-38 where the scale is also the same
If there is a work around for this then please let me know. Thanks!

hive> select cast('.0' as DECIMAL(1,1)), cast('0.0' as DECIMAL(1,1)), cast('0' as DECIMAL(1,1)), cast(0 as DECIMAL(1,1)), cast(0.0 as DECIMAL(1,1));
Query ID = xxxxxxx_20141024140000_e4dfdcc1-e4ad-4f84-bd48-198e29fd3757
Total jobs = 1
Launching Job 1 out of 1
Number of reduce tasks is set to 0 since there's no reduce operator
Starting Job = job_1413470329106_0052, Tracking URL = http://hdp8:8088/proxy/application_1413470329106_0052/
Kill Command = /usr/lib/hadoop/bin/hadoop job  -kill job_1413470329106_0052
Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 0
2014-10-24 14:01:10,256 Stage-1 map = 0%,  reduce = 0%
2014-10-24 14:01:27,644 Stage-1 map = 100%,  reduce = 0%, Cumulative CPU 6.51 sec
MapReduce Total cumulative CPU time: 6 seconds 510 msec
Ended Job = job_1413470329106_0052
MapReduce Jobs Launched: 
Job 0: Map: 1   Cumulative CPU: 6.51 sec   HDFS Read: 269 HDFS Write: 15 SUCCESS
Total MapReduce CPU Time Spent: 6 seconds 510 msec
OK
NULL    NULL    NULL    NULL    NULL
Time taken: 36.281 seconds, Fetched: 1 row(s)


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)